<!DOCTYPE html>
<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
  </head>
  <body>
    <p><span class="HwtZe" lang="en"><span class="jCAhz ChMk0b"><span
            class="ryNqvb">It is necessary to disable THP so often that
            I had to write a script for this purpose.</span></span> <span
          class="jCAhz ChMk0b"><span class="ryNqvb">Doing this using the
            service is much more convenient than doing it any other way,
            in my opinion.</span></span></span></p>
    <p><span class="HwtZe" lang="en"><span class="jCAhz ChMk0b"><span
            class="ryNqvb"><a class="moz-txt-link-freetext" href="https://github.com/yvoinov/memory-tools/blob/main/disable_thp.sh">https://github.com/yvoinov/memory-tools/blob/main/disable_thp.sh</a><br>
          </span></span></span></p>
    <div class="moz-cite-prefix">16.03.2024 14:23, Tarko Tikan via
      Unbound-users пишет:<br>
    </div>
    <blockquote type="cite"
      cite="mid:594fb6f6-ac15-40d1-bc23-0a69e4f86652@lanparty.ee">hey,
      <br>
      <br>
      <blockquote type="cite">The question is that stats metrics (mem*)
        are stable . They rapidly grow after startup and stay at a
        logical
        <br>
        maximum and don't keep growing.
        <br>
        <br>
        But the process size does.
        <br>
      </blockquote>
      <br>
      We saw the same on Debian due to THP like Oliver already
      mentioned. See <a class="moz-txt-link-freetext" href="https://github.com/NLnetLabs/unbound/issues/724">https://github.com/NLnetLabs/unbound/issues/724</a>
      <br>
      <br>
      We "solved" this by disabling THP on our unbound machines.
      <br>
      <br>
    </blockquote>
  </body>
</html>